Search papers, labs, and topics across Lattice.
Stanford University
1
0
2
6
Scaling LLMs significantly boosts social simulation accuracy in well-represented domains, but fails to enhance calibration for human cognitive biases.
